Trump Administration Pressures Big Tech for $15 Billion Energy Fund

The era of "cheap and infinite energy" that sustained Silicon Valley’s prosperity is coming to an end. As the explosion in power demand caused by the advancement of Artificial Intelligence (AI) reaches a critical point threatening national energy security, the U.S. government has finally taken action. Under the pretext of stabilizing electricity prices and preventing grid collapse, the Trump administration has begun pressuring Big Tech companies—including Google, Microsoft, and Amazon—to provide $15 billion (approximately 21 trillion KRW) in contributions for power generation facilities.

The End of the Grid 'Free-Ride': A $15 Billion Emergency Auction

This latest measure by the Trump administration goes beyond simple recommendations. The government is pushing a plan to mandate participation in an "Emergency Power Auction" to secure new power plant capacity, centered around the PJM Interconnection region, the nation’s largest grid operator. The logic is that since AI data centers are rapidly depleting grid reserves and driving up electricity bills for average households, the tech companies responsible for this demand should directly bear the cost of constructing new power plants.

According to investigations, the scale of this auction reaches approximately $15 billion. To cover these costs, data center operators must enter into long-term Power Purchase Agreements (PPAs) of 15 years or more with power generators. This event fundamentally shakes the cost structure of Big Tech. While companies previously operated under a variable Operating Expenditure (OpEx) model—paying only for the electricity they consumed—they are now being forced into a fixed Capital Expenditure (CapEx) structure, requiring them to directly build multi-billion dollar power plants or provide subsidies. The market is already reacting sensitively; news of the auction caused shares of GE Vernova, a power equipment company, to surge by more than 6%.

Green Pledges Clash with Baseload Realities

The sharpest point of this policy lies in the type of energy sources prioritized. For the sake of supply stability, the Trump administration is prioritizing the construction of "baseload" power plants, such as natural gas, coal, and nuclear power. This presents a critical dilemma for Big Tech companies that have placed RE100 compliance and carbon neutrality at the forefront of their marketing, championing 100% renewable energy use.

If these companies are forced by government pressure to fund fossil-fuel-based power plants, they could immediately be embroiled in "greenwashing" controversies. The sustainability brand value that Google and Microsoft have built over years is at risk of collapsing. Conversely, if they refuse the government's demands, they risk falling behind in power allocation priorities essential for AI data center expansion or facing astronomical surcharges on grid usage fees. Consequently, companies are in a position where they must either purchase more carbon credits to offset fossil fuel investments or accelerate investments in next-generation energy sources like Small Modular Reactors (SMRs).

The End of 'Data Center Alley' and Accelerated Vertical Integration

The government's hardline stance is also shifting data center location strategies. Traditionally, data centers were concentrated in hubs like Virginia, where grid infrastructure was well-established. Now, however, relying solely on the grid has become a risk in itself.

Companies are now pivoting toward "co-location" strategies, where power production and consumption occur in the same location. This involves building data centers directly adjacent to power plants or establishing "energy self-sufficient clusters" with their own generation facilities. This trend is accelerating the "vertical integration of energy," where tech companies acquire or directly operate energy firms. The fear that failing to control power supply risks could lead to losing leadership in AI is transforming these companies into infrastructure investors.

Practical Strategy: Energy Independence as AI Competitiveness

The strategy that data center operators and AI service providers must adopt at this point is clear:

First, diversify the energy portfolio. Companies must move away from relying solely on renewable energy and strategically adjust the proportion of gas- or nuclear-based PPAs. It is particularly important to preemptively secure partnerships with new baseload power plants in regions where regulatory approval is easier to obtain.

Second, redesign site selection. Companies must identify regions with rapid permitting processes and short grid connection wait times. According to JLL’s 2026 Global Data Center Outlook, demand for campus-style setups that integrate power generation and data centers is already exploding.

Third, design for extreme power efficiency. Beyond reducing hardware power consumption, a software-driven approach to adjusting model training loads based on power supply conditions is essential. Energy is no longer just a cost; it is the "most precious raw material" for AI operations.

FAQ

Q: Will participation in this $15 billion auction affect AI service pricing? A: Yes. Long-term PPAs of 15 years or more and multi-billion dollar infrastructure contributions mean a sharp increase in fixed costs. Companies are likely to raise prices for enterprise AI APIs or cloud services to offset these costs. As AI operating costs shift from a variable structure to an infrastructure investment structure, long-term upward pressure on costs is inevitable.

Q: Is it legally possible for companies that have declared RE100 to pay for fossil fuel power plant construction? A: Currently, this creates a conflict in implementation. It is highly likely that companies will either purchase large volumes of separate carbon credits to offset the fossil fuel portion of the electricity secured through the auction or take legal action to demand the government include renewable energy and Battery Energy Storage Systems (BESS) in the auction scope. These exceptions or detailed regulations are expected to be key points of contention during the final approval process by the Federal Energy Regulatory Commission (FERC).

Q: What is the impact on small and medium-sized data center operators? A: They face a significant crisis. The $15 billion auction is likely to be designed around Big Tech companies with massive capital. Smaller operators with insufficient funding may be pushed out of the competition for power or forced into situations where they must pay high rents to occupy power clusters established by Big Tech.

Conclusion

The era when tech companies remained solely in the world of software and algorithms is over. The Trump administration’s $15 billion auction pressure is forcing AI companies to wear a new badge: "Energy Infrastructure Investor." The future landscape of the AI industry over the next decade will be determined by the choices Big Tech makes between the public interest of grid stability and the corporate value of green sustainability. The winning edge in the AI race is now decided on power plant turbines rather than chip blueprints. Additional variables include the FERC’s schedule for regulatory revisions and the potential legal responses from the tech sector. The curtain has just risen on this massive power war.

Aionda